master thesis - universidade nova de lisboa arlindo tm 2015... · pom (>60 anos). endpoints do...
TRANSCRIPT
Master Thesis
The influence of menopausal status on bone turnover and
disease control in breast cancer patients with metastatic
bone disease treated with chemotherapy plus zoledronic
acid: an exploratory retrospective cohort study
Universidade Nova de Lisboa, Faculdade de Ciências Médicas
Northeastern University, College of Professional Studies
Arlindo Rebelo Ferreira
Medical Oncology Resident
Mentor: Luís Costa, MD, PhD
Lisbon - Boston 2015
Page 2 of 38
Abstract
Background: A recent meta-analysis showed that the adjuvant use of zoledronic acid (ZA) in
postmenopausal women with early breast cancer (BC) leads to a reduction in the risk of breast
cancer death by 17%. We investigated the effect of the hormonal status (pre [PrM] vs late post
menopause [PoM]) on bone turnover and disease control among women with BC and bone
metastases (BM) treated with ZA and chemotherapy (CT).
Methods: In this retrospective cohort study, we collected clinicopathologic variables, urinary N-
terminal telopeptide (NTX) and serum tumor marker levels from women with BC and BM treated
with CT and ZA. Patients were divided in PrM (<45 years) and PoM (>60 years). Study endpoints
were NTX, CA15.3 and CEA variation at 3, 6 and 9 months, and time to first-line CT failure and
survival. We performed multilevel mixed-effects linear regression models to assess the variation of
repeated measures and cox regression models for time to event outcomes.
Results: Forty patients were eligible for analysis (8 PrM and 32 PoM).
After introduction of ZA and CT, NTX and tumor markers declined in the overall cohort. Response
profile was similar between menopausal groups at month 3 and at later time points (p-value for
time-hormonal status interaction at month 3=0.957). Furthermore, tumor markers response profile
was also equal between groups.
Median time to first-line CT failure in PrM and PoM women was 15.2 and 17.4 months, respectively.
No significant difference between groups was found, either using a univariate analysis or after
controlling for visceral disease involvement (p=0.399 and 0.469, respectively). Likewise, no
differences in survival were found.
Conclusions: In this cohort, no differences were found in terms of NTX or tumor markers control
according to menopausal status. Similarly, no difference in time to first-line CT failure or survival
was found.
Keywords: breast cancer, bone metastases, zoledronic acid, bone remodeling markers, menopausal
status.
Page 3 of 38
Resumo
Introdução: Uma meta-análise recente demonstrou que uso adjuvante de ácido zoledrónico (AZ) em
mulheres pós-menopáusicas com cancro da mama precoce (CM) conduz a redução do risco de
morte por CM em 17%. Investigámos o efeito do estado hormonal (pré [PrM] vs pós-menopausa
tardia [PoM]) na remodelação óssea e controlo de doença em mulheres com CM e metástases
ósseas (MO) tratadas com AZ e quimioterapia (QT).
Métodos: Neste estudo de coorte retrospetivo, colhemos variáveis clinico-patológicas e
quantificámos o telopéptido N-terminal (NTX) urinário e marcadores tumorais (MT) séricos em
mulheres com CM e MO tratadas com QT e AZ. As doentes foram divididas em PrM (<45 anos) e
PoM (>60 anos). Endpoints do estudo: variação do NTX, CA15.3 e CEA nos meses 3, 6 e 9, tempo até
falência de QT de primeira-linha e sobrevivência. Quando apropriado foram usados os testes de
Wilcoxon rank-sum, modelo de efeitos lineares mistos, teste log-rank e modelo de Cox.
Resultados: Quarenta doentes foram elegíveis para análise (8 PrM e 32 PoM).
Depois da introdução de AZ e QT, os níveis de NTX e MT caíram no coorte global. O perfil de resposta
não diferiu entre grupos no mês 3 ou em tempos posteriores (valor-p para interação tempo-estado
hormonal no mês 3=0.957). Ademais, o perfil de resposta dos MT também não diferiu entre grupos.
O tempo mediano até falência de primeira-linha de QT em PrM e PoM foi de 15.2 e 17.4 meses,
respetivamente. Não foi identificada diferença significativa entre grupos, quer em análise univariada
quer após controlo para envolvimento visceral (p=0.399 e 0.469, respetivamente). Igualmente, não
houve diferenças em termos de sobrevivência.
Conclusões: Neste coorte, não foram identificadas diferenças no controlo de NTX ou MT em função
do estado menopausico. Igualmente, não foi identificada diferença no tempo até falência de
primeira-linha de CT ou sobrevivência.
Palavras-chave: cancro da mama, metástases ósseas, ácido zoledrónico, marcadores de
remodelação óssea, estado menopausico.
Page 4 of 38
List of figures and tables
Figures
Figure 1 - Interactions between bone and cancer cells in predominantly osteolytic breast cancer
lesions. Sites of action of denosumab and bisphosphonates are also noted. Bone metabolism with
resorption and formation occurs. The depicted mediators emphasize the predominant pathways.
Figure 2 – Study diagram.
Figure 3 – Mean (A), median (B) and mean log10 (C) variation of urine NTX at several time points
after diagnosis in the full cohort.
Figure 4 – Mean NTX at several time points after diagnosis according to menopausal status. Test of
differences in mean values is shown.
Figure 5 – Mean response profile of log10(NTX) at several time points after introduction of zoledronic
acid plus chemotherapy according to menopausal status. Test of interaction time - menopausal
group is shown.
Figure 6 – Mean response profile of CA15.3 and CEA at several time points after diagnosis in the full
cohort (A and B, respectively) and according to menopausal status (C and D, respectively).
Figure 7 – Mean response profile of log10(CA15.3) (A) and log10(CEA) (B) at several time points after
introduction of zoledronic acid plus chemotherapy according to menopausal status. Test of
interaction time - menopausal group is shown.
Figure 8 – Time to first-line chemotherapy failure according to menopausal status. Adjusted HR
controlling for visceral involvement is shown.
Figure 9 – Overall survival according to menopausal status. Adjusted HR controlling for HER2 status
is shown.
Tables
Table 1 – Types of bone remodeling markers.
Table 2 – Bone modifying agents.
Table 3 – Cohort baseline characteristics according to menopausal status.
Table 4 – Urine NTX at several time points from diagnosis.
Table 5 – Univariate association of baseline characteristics with time to first line chemotherapy
failure.
Table 6 – Univariate association of baseline characteristics with overall survival.
Page 5 of 38
Table of contents
Introduction ………………………………………………………………………………………………………… 6
Background and significance ……………………………………………………………………………….. 6
Study Objectives …………………………………………………………………………………………………. 14
Methods ……………………………………………………………………………………………………………… 15
Results ………………………………………………………………………………………………………………… 19
Discussion …………………………………………………………………………………………………………… 28
Conclusion ………………………………………………………………………………………………………….. 31
Acknowledgments ………………………………………………………………………………………………. 32
List of abbreviations ……………………………………………………………………………………………. 33
References ………………………………………………………….………………………………………………. 34
Page 6 of 38
Introduction This document constitutes the final thesis of the Double Master in Clinical Research from NOVA
Medical School (NOVA University) and Northeastern University.
This document is an original scientific dissertation (dissertação de natureza científica original) as per
Regulamento n.º 680/2010.
This project was presented at the “Spring Meetings 2014” (Évora, Portugal) in poster format and
was awarded a prize for the second best poster presented in the meeting.
Background and significance
a) Breast cancer epidemiology
Breast cancer (BC) is the most frequently diagnosed and the most common cause of women cancer
related death in Europe with an estimated incidence and death rate of 108.8 and 22.4 cases per 100
000 women per year, respectively.[1] More specifically, the estimated incidence and death rate in
Portugal is of 85.6 and 18.4 cases per 100 000 women per year, respectively.[1] Furthermore, breast
cancer is also the most common cause of cancer related death in women both in developed and
developing regions.[2]
In Europe and in the USA, the majority of patients currently diagnosed with BC present with early
or locally advanced disease in terms of stages of disease progression, allowing the use of therapies
with curative intent.[3] However, 5-10% of breast cancers are diagnosed with metastatic disease.
Furthermore, from those treated with curative intent, around 40% will develop disease
recurrence.[4] While locoregional recurrence might be eligible for curative interventions, those with
distant relapse are almost universally incurable.
b) Bone metastases and skeletal related events
In patients with metastatic breast cancer bone is the most common involved site, either as de novo
disease or after disease relapse. While the 5-years incidence of bone metastases in patients treated
with curative intent is around 5%, 70% of those with metastatic disease have bone involvement at
some time during disease evolution.[5] Liver, lung and less frequently the brain are also typical
target organs of breast cancer metastatic spread.
The exact mechanisms of breast cancer cells tropism to the bone is not fully understood. However,
both anatomical characteristics, as blood drainage following the Batson plexus, and molecular
features, as e.g. the chemoattractant action of bone marrow derived CXCL12 that induces tropism
of tumor cells expressing CXCR4 and CXCR7, partially explain the homing process of breast cancer
cells to the bone. After reaching the bone, Src activity seems to play a major role in supporting
Page 7 of 38
cancer cell survival in the bone marrow microenvironment. This process appears to be accomplished
by facilitating CXCL12-CXCR4-AKT signaling and by conferring tumor cells resistance to TNF-related
apoptosis-inducing ligand (TRAIL).[6] Bone metastasis are also more common in patients whose
tumors express the estrogen receptor (ER).[7] As a matter of fact, Src is active in most estrogen
receptor positive and in a minority of estrogen receptor negative tumors.[6] Furthermore, Src and
the ER seem to interact.[8] Despite this interaction, Src expression signature is more strongly
associated with bone disease than the ER per se, an association that seems independent in nature
from the ER. In line with this, even when restricting to patients with ER-negative tumors, Src
expression is still highly associated with bone metastatic development, a fact that underlines the
relevance of Src expression as, at least, a marker of activation of relevant pathways leading to a
successful adaptation of cancer cells to bone environment.[6]
Breast cancer cells further elicit a cascade of
events that lead to a profound change in bone
metabolism. Bone is under continuous
remodeling, a process accomplished by the
coupled activity of osteoblasts (bone forming
cells) and osteoclasts (bone resorbing cells).
Cancer cells reshape this equilibrium to
activate osteoblasts. This process is at least
partially accomplished through the production
of parathyroid hormone-related peptide
(PTHrp), tumor necrosis factor α (TNF-α),
interleukin 1 (IL-1), IL-6, IL-8 and IL-11 by
cancer cells.[9] Activated osteoblasts release
activator of nuclear factor κ B ligand (RANKL)
that ultimately activates osteoclasts; hence, it
induces bone resorption, which leads to the
release of growth factors entrapped in the
bone matrix (transforming growth factor-β
(TGF- β), bone morphogenetic proteins
(BMPs), insulin like growth factor (IGF) and
fibroblast growth factor).[9, 10] This last step
is the closing gap in the loop, as cancer cells will
benefit from these growth stimulation factors,
therefore leading to more bone turnover
activation: a loop referred as the vicious cycle
of bone metastases.[11]
Bone remodeling modifications ultimately
conduct to increased bone fragility[12] and
development of skeletal related events (SRE).
Figure 1 - Interactions between bone and cancer cells
in predominantly osteolytic breast cancer lesions.
Sites of action of denosumab and bisphosphonates
are also noted. Bone metabolism with resorption and
formation occurs. The depicted mediators emphasize
the predominant pathways. See text for further
details.
Page 8 of 38
Skeletal related events include a series of adverse outcomes derived from the presence of bone
metastases, namely excessive pain requiring intervention, as radiotherapy or surgery, bone fracture,
myeloid compression and hypercalcemia. The interaction between cancer and bone cells can induce
the development of several types of lesions: lytic (mostly bone degrading), blastic (mostly bone
forming) or mixed (lesions presenting a mixture of lytic and blastic features). Either of these lesions
represent a profound change in bone architecture and therefore of bone resistance to mechanical
stress. In patients without therapy aiming to control bone metabolism, a rate of approximately 4
SREs per year is expected, significantly impairing patients’ quality of life and ultimately, survival.[13,
14]
c) Markers of bone remodeling
Bone remodeling involves degradation of current bone matrix and formation of new bone matrix;
in fact, both biochemical processes lead to the release of several biochemical by-products amenable
of being quantified in, e.g., blood or urine.[15] NTX (amino-terminal crosslinked telopeptide of
collagen type I) is a by-product of collagen degradation and one of the most commonly used markers
of bone metabolism, in specific of the bone resorbing arm. On the other hand, ALP (alkaline
phosphatase) is a by-product of osteoblast activity; hence, a marker of bone formation.[15] Other
types of bone remodeling markers are listed in table 1.
Bone markers are relatively non-invasive indicators of bone metabolism and several studies tested
their role in the clinical setting. Urinary NTX has been characterized in patients with or without bone
metastases, and receiving or not bone modifying agents.[16, 17] In fact, this and other bone
remodeling markers (as CTX, carboxy-terminal crosslinked telopeptide of collagen type I) are being
explored in several perspectives in the clinical setting, namely a) to evaluate the risk of developing
bone metastases[18], b) to diagnose new bone metastases[19, 20], c) to evaluate response to
therapy/patient outcomes (SRE, bone disease progression and overall survival)[21, 22], as well as d)
as a tool to titrate the dose and frequency of administration of bone targeted agents[23]. The most
relevant use of bone remodeling markers was however during the trials that led to the approval of
new bone targeted agents as denosumab, when the variation of these markers was used as a central
study endpoint.[24] Noteworthy, its role in the clinical setting is still exploratory and, therefore, not
recommended outside clinical trials.[25]
Page 9 of 38
Table 1 – Types of bone remodeling markers.
Bone formation markers
Bone specific alkaline phosphatase (BALP)
Osteocalcin (OC)
Propeptides of type I procollagen (P1NP and P1CP)
Bone resorption markers
Amino-terminal crosslinked telopeptide of collagen type I (NTX)
Calcium
Carboxy-terminal crosslinked telopeptide of type I collagen (ICTP or CTX-MMP)
Carboxy-terminal crosslinked telopeptide of collagen type I (CTX)
Deoxypyridoline (DPD)
Hydroxyproline (Hyp), Hydroxylysine (Hyl)
Non collagenous bone matrix proteins: bone sialoprotein (BSP) and osteopontin (OP)
Osteoclast-derived enzymes: tartarate resistant acid phosphatase 5b (TRAP5b) and cathepsin k and L
Pyridinoline (PYD)
Parathyroid hormone related peptide (PTH-rp)
Vitamin D
Regulators of bone turnover
Dickkopf-1 (DKK-1)
Osteoprotegerin (OPG)
Receptor activator of NFƙB ligand (RANKL)
Sclerostin
d) Bone modifying agents
After the establishment of metastatic bone lesions two strategies are commonly implemented to
control the impact of tumor cells in the bone: the use of anti-tumor treatments and/or bone
directed treatments (osteoclast inhibitors). Anti-tumor treatments include targeted therapy (e.g.,
hormone therapy or anti-HER therapies), chemotherapy or radiation therapy. On the other hand,
bone directed therapy is composed of bone modifying agents, namely bisphosphonates (BP) and
denosumab (table 2).
Page 10 of 38
Table 2 – Bone modifying agents. NA – not applicable.
Pharmacologic Category Agents Relative potency
Bisphosphonates
Non nitrogenous-bisphosphonates
Etidronate
Clodronate
Tiludronate
1
10
10
Nitrogenous-bisphosphonates
Pamidronate
Alendronate
Ibandronate
Risedronate
Zoledronate
100
500
1000
2000
10000
Monoclonal antibodies Denosumab NA
BP are a group of pyrophosphate analogous with an intense tropism to the bone where, after
incorporation into the matrix and subsequent absorption by osteoclasts (during the process of bone
resorption), they act as potent osteoclast inhibitors (figure 1). There are two classes of BP, non-
nitrogenous and nitrogenous. While the first act by substituting the terminal pyrophosphate moiety
of adenosine triphosphate (ATP), resulting in a nonfunctional ATP molecule that competes for
functional ATP in the osteoclast energy metabolism, the latter inhibit the osteoclast enzyme farnesyl
diphosphate synthase (FDS). FDS is a critical isoprenyl diphosphate synthase or prenyl transferase
in the mevalonate pathway that is fundamental for the synthesis of sterol isoprenoids, such as
cholesterol, and non-sterol isoprenoids, such as dolichol, heme-A, isopentenyl tRNA and
ubiquinone.[26] The inhibition of isoprenoid – farnesyl diphosphate and geranylgeranyl
diphosphate synthesis by the inhibition of FDS prevents the post-translational prenylation of small
GTPase signaling proteins such as Rho, Rac, Rab, Rap and Ras leading to disruption of osteoclast
function and induction of apoptosis.[26]
On the other hand, the fully humanized monoclonal antibody denosumab acts by sequestering
RANKL (figure 1). This process limits the natural action of RANKL on the osteoclast precursor surface
receptor RANK, preventing osteoclast differentiation and ultimately bone resorption.
BP (in specific zoledronic acid, pamidronate and ibandronate), but also denosumab, are widely used
in clinical practice as tools to reduce the morbidity of bone metastases. Despite subtle differences
in relative efficacy of this agents[27, 28], major international guidelines support the use of either
zoledronic acid (ZA) or denosumab to prevent and delay SREs.[25, 29]
Page 11 of 38
e) Anti-tumoral action of bisphosphonates
BP are critical tools to prevent and delay SREs; however, pre-clinical and clinical studies are
increasingly supporting a direct and indirect anti-tumoral action of these agents. Some of the anti-
tumor mechanisms of action of BP include a direct cytotoxic action, but also other indirect
mechanisms, such as anti-angiogenesis and immune system activation.[30] While the preclinical
identification of this effects seems clear, the clinical translation of these findings is less clear, and,
so far, seems to be restricted to hormone-sensitive patients under an estrogen depleted
environment (either due to post-menopausal status or due to ovarian suppression). Furthermore,
the effect is more clear in nitrogenous-bisphosphonates, in specific for ZA.[31] Of note, more than
80% of the women with BC are post-menopausal, which opens a wide window for eventual
intervention/applicability.[3] Nevertheless, at present, available clinical practice guidelines do not
recommend the use of BP as a therapeutic tool in the adjuvant treatment of breast cancer.[25, 29,
32]
In the last decade, a group of trials documented relevant pieces of information regarding the anti-
cancer activity of BPs in the adjuvant treatment of breast cancer. The ABCSG-12 was a phase III trial
that tested the adjuvant use of ZA (4 mg every 6 month for 3 years) in 1803 pre-menopausal women
under ovarian suppression and tamoxifen or anastrazol.[33, 34] After a median follow-up of 8 years,
those patients receiving ZA had a 33% decrease in the risk of recurrence (Hazard-ratio [HR] for
disease-free survival [DFS] 0.77, 95% Confidence Interval [CI] 0.60 – 0.99; p=0.042) and a trend
towards a decrease in the risk of death (HR for overall survival [OS] 0.66, 95% CI 0.43-1.02; p=0.064).
Noteworthy, this effect was clearer in patients older than 40 years of age (HR for DFS 0.70, 95% CI
0.51 – 0.96 vs. 0.95, 95% CI 0.62 – 1.46 in women with less than 40 years of age; HR for OS 0.56,
95% CI 0.31 – 0.96 vs. 0.94, 95% CI 0.45 – 1.98 in women with less than 40 years of age).
The ZO-FAST and Z-FAST trials also tested the effect of BPs in terms of cancer related outcomes.
Despite theirs design to test as central outcomes bone health related outcomes (as variation in bone
mineral density and the occurrence of bone fracture), they also tested for cancer related outcomes
as secondary outcomes (disease recurrence). In ZO-FAST (1065 women) and Z-FAST (602 women)
postmenopausal women receiving adjuvant letrozole were randomized for upfront zoledronic acid
(4 mg every 6 months) or zoledronic acid only after bone fracture or significant decline in bone
mineral density.[35, 36] Despite de concordant results between trials in terms of improved bone
mineral density for patients receiving upfront ZA, only those patients enrolled in the ZO-FAST
derived a benefit in terms of disease recurrence (HR for DFS 0.66, 95% CI 0.44 – 0.97).[35, 36]
The AZURE trial was designed to have as a primary outcome the clinical impact of adjuvant ZA in
terms of breast cancer control related outcomes.[37] In this phase III study, 3360 pre or post-
menopausal women were randomized to standard adjuvant therapy with or without ZA (4 mg every
3-4 weeks for the first 6 months and then every 3-6 months for a total of five years). Primary
outcome was disease-free survival. At a median follow-up of around 5 years, no significant
difference in terms of DFS or OS was documented in the overall cohort. However, in a pre-planned
subgroup analysis, late post-menopausal women (those > than 5 years post-menopause) derived a
Page 12 of 38
significant benefit: 25% reduction in the risk of invasive disease or death (adjusted-HR 0.75, 95% CI
0.59-0.96) and 26% reduction in the risk of death (adjusted-HR 0.74, 95% CI 0.55-0.98).
As referred before, most of the available clinical evidence of the anti-tumor action of BPs is derived
from trials that used zoledronic acid; however, other BPs were also tested. The NSABP B-34 tested
the role of clodronate both as a strategy to test the action of this BP in terms of bone health related
outcomes (as bone mineral density), but also in terms of disease recurrence.[38] In this study, 3323
women (pre and post-menopausal) were randomized to receive standard adjuvant therapy with or
without oral clodronate (1600 mg/d for a total of 2 years). After a median follow-up of around 7
years, only those 50 years of age or older had a benefit in terms of DFS (bone and extra-bone).
However, no clear improvement in terms of OS was documented for both age groups. Pamidronate
was also tested as a strategy to reduce the risk of disease recurrence in breast cancer.[39, 40] In a
small trial Kokufu and colleagues recruited 90 patients with node positive (≥4 positive nodes) early
breast cancer treated with curative intention to receive pamidronate (n=33; 45 mg pamidronate 4
times every 2 weeks) or standard follow-up (n=57) based on personal preference.[39] After a
median follow-up of five years, a reduction in the incidence of bone metastases in the group
receiving pamidronate was found (12.1% vs. 40.4% in the control group; p=0.005). A non-significant
numerical difference favoring pamidronate was also found in terms of distant metastases (36.4 vs.
56.1%; p=0.071) and non-osseous metastases (33.3 vs. 52.6%, p=0.077). Nevertheless, no difference
regarding overall survival or disease-free survival was found. Another study testing a similar
hypothesis recruited 429 perimenopausal stage I-III breast cancer patients receiving therapy with
curative intent.[40] From these, 258 received pamidronate (15mg iv every 4 weeks and oral
pamidronate 100mg daily during adjuvant chemotherapy) while 171 didn’t. Even though patients
receiving pamidronate had a lower incidence of bone metastases (2.3% vs. 8.7% in the control
group; time of endpoint evaluation not specified) no difference was found in terms of disease-
free survival or overall survival.
Finally, a recent meta-analysis analyzed individual patient data from several randomized trials
comparing de use of BP to no BP (either placebo-controlled or open control) in terms of the risk of
recurrence and death.[31] In this study (preliminary data after analysis of 75% of 23,573 patients of
interest), a statistically significant reduction of 17% in the risk of death and distant recurrence was
found for post-menopausal women treated with BP only. Furthermore, a reduction of 35% in the
risk of bone-specific recurrence was shown, an effect also restricted to the group of post-
menopausal women. As a matter of fact, pre-menopausal women seem not to benefit from BPs in
terms of the risk of disease recurrence and death.
When taken together, an increasing body of clinical evidence is pointing towards an effect of BPs,
and more specifically ZA, as a tool to reduce the risk of recurrence and death when used in the
adjuvant treatment of post-menopausal women (5 years after menopause establishment or older
than 40 under ovarian suppression) with hormone receptor positive breast cancer. We should note
however that no current international clinical guideline is recommending the use of BPs in this
setting.[25, 29, 32]
Page 13 of 38
f) Framing the research question
Metastatic breast cancer survival has improved over time, mostly after the introduction of most
effective therapies, as taxanes, aromatase inhibitors and anti-HER agents. Other relevant gains were
achieved with more rational use of already available drugs tailored to certain risk specific groups.
BPs, and most specifically ZA, seem to have antitumor properties which depend on the systemic
hormonal environment. In fact, only late post-menopausal women (or premenopausal under
ovarian suppression therapy) seem to benefit from zoledronic acid in terms of disease recurrence
and survival. However, the available data are only derived from patients in the early breast cancer
setting, and there is no consistent evidence that BPs improve survival in metastatic breast cancer
patients nor was the effect of the hormonal environment tested in this circumstance.
In this study we aim to test if the hormonal environment significantly impacts patient outcomes, as
bone disease control, by means of the measurement of urinary NTX levels, serum levels of tumor
markers, time to first-line chemotherapy failure and overall survival in patients with metastatic
breast cancer affecting the bone and being treated with zoledronic acid and chemotherapy.
Page 14 of 38
Study Objectives To test whether the hormonal environment (pre vs. late post-menopausal) significantly impacts
bone disease control, as measured by urinary levels of NTX, serum levels of tumor markers, time to
first-line of chemotherapy failure and overall survival in patients with breast cancer affecting the
bone and being treated with zoledronic acid and chemotherapy.
a) Primary objective: To determine the impact of the hormonal environment (pre vs. post-
menopausal) on urinary levels of NTX at 3 months in patients receiving zoledronic acid and
chemotherapy;
Hypothesis: Post-menopausal patients will present a more pronounced decline in the
urinary NTX levels at 3 months when compared to pre-menopausal women.
b) Secondary objectives: To determine the impact of the hormonal environment (pre vs. post-
menopausal) in patients receiving zoledronic acid and chemotherapy on the following
parameters:
Urinary NTX levels at 6 and 9 months;
Hypothesis: Post-menopausal patients will present an increased decline in the urinary levels
of NTX at 6 and 9 months when compared to pre-menopausal women.
Serum CA15.3 and CEA levels (tumor markers);
Hypothesis: Post-menopausal patients will present an increased decline in the serum levels
of CA15.3 and CEA when compared to pre-menopausal women.
Time to first line chemotherapy failure;
Hypothesis: Post-menopausal patients will present a longer time to first line chemotherapy
failure when compared to pre-menopausal women.
Overall survival;
Hypothesis: Post-menopausal patients will present a longer overall survival when compared
to pre-menopausal women.
Page 15 of 38
Methods a) Study population
Patients with the diagnosis of metastatic breast cancer involving the bone (de novo or post-
disease recurrence), receiving zoledronic acid and chemotherapy, diagnosed and treated at
the Department of Medical Oncology from Hospital de Santa Maria – CHLN from 2000 and
2009 and with quantifications of urinary NTX available.
Evidence of breast cancer was obtained from the histopathology report, while evidence of
bone involvement was obtained from imaging.
We will exclude patients with previous diagnosis of other oncologic disease, except
contralateral breast cancer that is disease-free for more than 5 years, non-melanoma skin
cancer or pre-invasive cervix cancer. We will also exclude patients with a follow-up of less
than 3 months, with previous treatment for metastatic disease, previous treatment with
BPs and pregnant women or women with psychiatric conditions.
b) Study design and diagram
This study is a single center retrospective cohort study.
Figure 2 – Study diagram.
Women with Metastatic breast cancer
Evaluation of eligibility
Treatment with first line chemotherapy + zoledronic acid
Pre-menopausal women Post-menopausal women
Page 16 of 38
c) Data collected and study procedures
To collect the clinical variables of interest we directly extracted clinical information from medical
records, both physical and electronic.
The collected variables were the following:
Demographics: age at diagnosis, race and menopausal status;
Disease and tumor characterization: date of diagnosis, WHO performance status, AJCC
staging (T, N, number of involved lymph nodes and M), hormone receptor status, HER2
status, grade, histology, previous (neo)adjuvant chemotherapy regimen, (neo)adjuvant
trastuzumab, (neo)adjuvant chemotherapy, date of disease relapse, metastatic disease
characterization (date of metastatic diagnosis, number and localization of lesions);
Biomarkers and imaging: NTX, CEA, CA15.3, CT (thorax/abdomen/pelvis), scintigraphy,
simple radiography of the bone.
Other endpoints: date of disease progression (visceral and bone), date of introduction of
first and second line chemotherapy, date of SRE and date of death.
d) Definition of outcomes and other definitions
Biomarkers response: variation of NTX and tumor markers (CEA and CA15.3) from baseline
(before introduction of BPs) and time points of interest.
Time to first line chemotherapy regimen failure: time from introduction of BPs plus first line
chemotherapy to the introduction of second line chemotherapy regimen.
Menopausal status: Menopausal status will be assigned by age group. Women will be assigned
as pre-menopausal if less than 45 years of age; late post-menopausal women will be those with
more than 60 years of age.
e) Evaluation of urinary NTX
We used Osteomark NTx Urine ELISA (Inverness) to determine urinary NTX. The detection limit
of the NTX assay is 20 nM BCE (bone collagen equivalents units – assay value, does not include
creatinine excretion). The urinary levels of NTX in bone collagen equivalent units are expressed
as the ratio of NTX to urine creatinine excretion. Urinary NTX was defined as low (NTX<50 nmol
BCE/mmol creatinine) or high (NTX ≥ 50 nmol BCE /mmol creatinine). The cutoff values for NTX
were chosen to reflect the approximated levels of normal. However, the normal range for
urinary NTX varies according to several factors, as age, gender and endocrine environment.
While the upper level of normal (ULN) in young healthy adults is approximately 50 nmol/mmol
creatinine, after menopause, the ULN level is approximately 100 nmol BCE/mmol creatinine.[22]
Page 17 of 38
f) Statistical considerations
Descriptive statistics
Tabulation of baseline demographic, clinical, pathological and treatment characteristics
were performed according to menopausal status using Pearson’s χ2, Fisher’s exact test, t-
test or Wilcoxon rank-sum test when appropriate. Graphical representations of variation of
NTX and tumor markers through time were performed.
Analysis of outcomes
1) NTX and tumor markers (CEA and CA15.3) at baseline, 3, 6 and 9 months in the
full cohort and by menopausal status
a. Wilcoxon rank-sum test was performed to test differences between
menopausal groups at specific time points: baseline, 3, 6 and 9 months.
b. Linear mixed effects models for longitudinal data were used to model
mean variation of NTX and tumor markers over time given the
correlated nature of the outcomes (repeated measures). We used a
model with random intercepts and slopes based on the hypothesis that
individuals vary not only in their baseline level of response, but also in
terms of their changes in the mean response over time. We further
allowed the mean of intercepts and slopes to depend on the fixed effect
menopausal status category (independent variable of primary interest).
While the fixed effect parameter menopausal status informs us on how
the sample means differ between pre and post-menopausal women,
the random effect parameters represent the broad variability among
subjects. We restricted analysis to 3 time intervals, from baseline to 3,
3 to 6 and 6 to 9 months.
c. NTX and tumor markers were analyzed in absolute and log transformed
values as a strategy to focus on relative changes of biomarker at
different time points and obtain a normal distribution of biomarker
levels. Main results will be presented using the log transformed version
of the biomarkers level.
2) Time to first line chemotherapy failure and overall survival
a. The time to first line chemotherapy failure was analyzed using survival
analysis techniques. The events of interest were failure of first line
chemotherapy and death. We used the Kaplan-Meier estimator to
estimate the survival function and Kaplan-Meier plot according to
menopausal status.
b. Log-rank test was used to test the univariate difference between
survival curves of the group of pre and post-menopausal women.
Page 18 of 38
c. Cox proportional hazards model was used to test the multivariate
difference between survival curves according to menopausal status.
Outline of the model:
Outcome/dependent variable: failure of first line chemotherapy
and death;
Independent variable: menopausal status;
Covariates: estrogen receptor status, HER2 receptor status,
histologic grade, histology and concomitant visceral involvement.
Due to limitations of sample size/number of events only bivariate
models were built.
Sample size calculation
This study is of exploratory in nature, so no prior knowledge is available in terms of primary
outcome to perform a sample size calculation. Furthermore, we are constrained by a limited
set of NTX evaluations. No further NTX evaluations were performed due to funding
constrains. We will use the present results as the foundation for future assessment of
feasibility of a larger study testing the same hypothesis.
Sources of confounding and strategies to avoid confounding
The following variables were selected as potential sources of confounding: estrogen
receptor status, HER2 receptor status, histologic grade, histology and concomitant visceral
involvement. Multivariate analysis used as a strategy to overcome confounding.
g) Ethical considerations
This study was approved by the ethics committee of the Lisbon Academic Medical Center. A
waiver of the use of informed consent was granted based on the fact that urine and blood
samples had already been collected for the purpose of the evaluation of NTX and tumor markers
in the context of related studies and clinical practice, respectively. Moreover, the investigation
per se included no more than minimal risk in the form of loss of confidentiality. To overcome
this liability, several measures were taken. Some of these measures included the de-
identification of patients and subsequent re-identification using a study specific numerical study
code (p.e. 001); variables were also coded numerically (e.g., yes/no variables will be coded as
1/0, respectively) and a separated code of variables created. Moreover, electronic files were
always kept inside CHLN network or in a very limited set of other encrypted computers that had
continuous anti-virus protection. Furthermore, electronic files were always stored using an
alphanumeric password protection (either the specific file or after placing the file inside a
password protected .rar library).
Page 19 of 38
Results a) Cohort characteristics
A total of 40 women were included in this study, 8 (20.0%) of which pre-menopausal and 32 (80.0%)
post-menopausal. Relevant patient’s clinical and pathologic characteristics are summarized in table
3. As a matter of fact, the groups were reasonably balanced regarding most of the selected variables.
Noteworthy, median ages are compatible with corresponding menopausal status. Furthermore,
most tumors were estrogen receptor positive, as expected from a cohort of patients with bone
disease as first site of relapse, and had a fairly good performance status, which enabled the use of
chemotherapy as a treatment option.
Table 3 – Cohort baseline characteristics according to menopausal status.
Pre-menopausal women
Late post-menopausal women
P-value
Number, % 8 (20.0) 32 (80.0) -
Age at diagnosis, median (IQR) 39.8 (35.9 – 44.2) 65.6 (60.1 – 67.8) -
Age at diagnosis of metastatic disease, median (IQR)
44.8 (40.6 – 46.5)
67.2 (62.9 – 72.9)
-
Disease-free interval (excludes stage IV patients), median (IQR)
39.7 (22.9 – 85.8)
42.6 (28.0 – 83.4)
0.366
Histology, n (%) Invasive Ductal carcinoma
Other Unknown
7 (100)
0 (0) 1 (12.5)
19 (79.2) 5 (20.8) 8 (25.0)
0.737
Estrogen receptor status Positive
Negative Unknown
7 (100)
0 (0) 1 (12.5)
21 (80.8) 5 (19.2) 6 (18.8)
0.559
HER2 status Positive
Negative Unknown
1 (20.0) 4 (80.0) 3 (37.5)
5 (27.8)
13 (72.2) 14 (43.8)
0.608
Stage IV (at diagnosis), n (%) Unknown
1 (12.5) 2 (25.0)
9 (47.4) 13 (40.6)
0.345
Visceral involvement at diagnosis of metastatic disease, n (%)
Unknown
3 (37.5)
0 (0)
4 (15.4) 6 (18.8)
0.315
Performance status, median (IQR) 0 (0 - 0) 0 (0 - 1) 0.132
Total follow-up, median (IQR) (months)
36.8 (30.0 – 53.0) 53.3 (24.9 – 62.7) 0.080
Page 20 of 38
a) Urinary NTX
The majority of patients had elevated levels of urinary NTX at baseline, which is compatible with
active disease in the bone. Moreover, when evaluating the full cohort, after the introduction of
zoledronic acid and chemotherapy (at diagnosis - baseline) a decrease of NTX is notable (table 4 and
figure 2), with 41.2% of the patients with elevated NTX at baseline reaching an NTX < 50 nmol/mmol
creatinine at month 3. Furthermore, after log-transforming NTX values, a trend for an overall decline
in mean log10(NTX) in the interval 0 – 3 months (p=0.125) and a statistically significant decline of
log10(NTX) in the interval 3 – 6 months (0.013) was found. Importantly, the proportion of patients
that were lost to follow-up in terms of urinary NTX measurements was high: while at month 6 75%
of the patients had a valid measurement of urinary NTX, at month 9 only 37.5% had. Of note, no
clear differences in terms of age, disease progression or vital status were found between lost to
follow-up and retained patients; therefore, we considered this missing information as missing
completely at random.
Table 4 – Urine NTX at several time points from diagnosis. NTX measured in nmol/mmol creatinine.
Baseline 3 months 6 months 9 months Number, % 40 (100) 39 (97.5) 30 (75.0) 15 (37.5)
Urine NTX, median (IQR) All
Pre-menopausal Post-menopausal
119.8 (66.5 – 211.1) 70.9 (43.5 – 137.2)
121.3 (71.9 – 245.0)
58.2 (25.6 – 135.9) 49.7 (32.9 – 80.0)
61.0 (25.6 – 161.9)
46.23 (20.7 – 123.9)
26.2 (12.3 – 47.2) 55.8 (24.8 – 124.8)
62.1 (32.8 – 162.9)
100.3 (50.8 – 200.7) 57.4 (31.5 – 139.3)
Elevated urine NTX, n (%) All
Pre-menopausal Post-menopausal
34 (85.0) 5 (62.5)
29 (90.6)
23 (57.5) 4 (50.0)
19 (59.4)
24 (60) 3 (37.5)
21 (65.6)
35 (87.5) 8 (100)
27 (84.4)
NTX normalization (if elevated at baseline), n (%)
All Pre-menopausal
Post-menopausal
- - -
14 (41.2) 4 (80.0)
10 (34.5)
14 (41.2) 5 (100) 9 (31.0)
4 (11.8) 0 (0)
4 (13.8)
Page 21 of 38
Figure 3 – Mean (A), median (B) and mean log10 (C) variation of urine NTX at several time points
after diagnosis in the full cohort.
When stratified by menopausal status (figure 3), post-menopausal women present persistently
absolute higher levels of mean NTX, namely at baseline and at 3 months; however, such differences
do not reach statistical significance. In fact, such differences are non-significant through all the
follow-up period. After log-transforming NTX values (figure 4), the early response (baseline to month
3) to zoledronic acid and chemotherapy is equal between groups (p=0.957). Both groups seem to
“escape” zoledronic acid plus chemotherapy control after month 6, as demonstrated by an increase
of NTX in both groups. Of note, in absolute and relative terms, urinary NTX levels seems to increase
more pronouncedly in the group of premenopausal women in this interval (figure 3 and 4); however,
this difference is non-significant (p=0.262). Moreover, when all the follow-up period is taken
together, the response profile of urinary NTX levels to zoledronic acid and chemotherapy does not
differ between pre and post-menopausal women (p=0.314).
A B
C
-100
01
00
20
03
00
40
0M
ean
NT
X
0 3 6 9Months
NTX Standard deviation
05
01
00
15
02
00
Me
dia
n N
TX
0 3 6 9Months
NTX Interquartile range
11
.52
2.5
Me
an
log
(NT
X)
0 3 6 9Months
log(NTX) Standard deviation
Page 22 of 38
11
.52
2.5
Me
an
log
(NT
X)
0 3 6 9Months
Premenopausal Postmenopausal
Standard deviation
Figure 4 – Mean NTX at several time points after diagnosis according to menopausal status. Test of
differences in mean values is shown.
Figure 5 – Mean response profile of log10(NTX) at several time points after introduction of zoledronic
acid plus chemotherapy according to menopausal status. Test of interaction time - menopausal
group is shown.
-10
00
100
200
30
04
00
Me
an
NT
X
0 3 6 9Months
Premenopausal Postmenopausal
Standard deviation
p = 0.957 p = 0.316 p = 0.262
p = 0.314
p = 0.187 p = 0.444 p = 0.120 p = 0.471
Page 23 of 38
b) Tumor markers (CA15.3 and CEA)
The majority of patients had elevated levels of serum CA15.3 and CEA at baseline, which is
compatible with active breast cancer. After the introduction of zoledronic acid and chemotherapy
(at diagnosis - baseline) a slight numerical decrease of both markers is identifiable (figure 5 and
figure 6); however, only a trend for decreased values through the interval 0 – 9 months is found (p
= 0.058 and 0.057, for CA15.3 and CEA respectively).
When stratified by menopausal status (figure 6), no clear differences are found between pre and
post-menopausal women in the log transformed values, either at baseline or at the following time
points, both for CA15.3 and CEA.
Figure 6 – Mean response profile of CA15.3 and CEA at several time points after diagnosis in the full
cohort (A and B, respectively) and according to menopausal status (C and D, respectively).
-200
02
00
40
06
00
80
0M
ean
CA
15
.3
0 3 6 9Months
Premenopausal Postmenopausal
Standard deviation
A B
C D
-200
02
00
40
06
00
Me
an
CA
15
.3
0 3 6 9Months
CA15.3 Standard deviation
-50
05
01
00
Me
an
CE
A
0 3 6 9Months
CEA Standard deviation
-50
05
01
00
Me
an
CE
A
0 3 6 9Months
Premenopausal Postmenopausal
Standard deviation
Page 24 of 38
Figure 7 – Mean response profile of log10(CA15.3) (A) and log10(CEA) (B) at several time points after
introduction of zoledronic acid plus chemotherapy according to menopausal status. Test of
interaction time - menopausal group is shown.
0.5
11
.5M
ean log
(CE
A)
0 3 6 9Months
Premenopausal Postmenopausal
Standard deviation
11
.52
2.5
3M
ean log
(CA
15.3
)
0 3 6 9Months
Premenopausal Postmenopausal
Standard deviation
p = 0.997 p = 0.402 p = 0.531
p = 0.762
p = 0.683 p = 0.670 p = 0.315
p = 0.478
A
B
Page 25 of 38
0
10
20
30
40
50
60
70
80
90
100
Sw
itch
ing fir
st lin
e c
he
mo
the
rap
y(%
)
17 17 11 9 7 5Post-menopause8 6 6 5 4 1Pre-menopause
No. at risk
0 5 10 15 20 25Follow-up time (months)
Pre-menopause Post-menopause
c) Time to first-line chemotherapy failure
When evaluating time to first-line chemotherapy failure, the full cohort presented a median follow-
up time of approximately 17 months (IQR 6.3 – 23.2). A total of 30 events (90.9%) were documented,
of which 8 (100%) in the premenopausal group and 22 (88.0%) in the post-menopausal group. The
median time to chemotherapy failure was of 15.3 months for pre-menopausal women and 17.4 for
post-menopausal women. In figure 7 a representation of the survival function according to
menopausal status is depicted. When evaluating the univariate association of menopausal status
and other clinical and pathological variables with change of first line chemotherapy, only the
presence of visceral involvement at diagnosis of metastatic disease was significantly associated with
time to chemotherapy switch (table 5). Furthermore, when testing the role of menopausal status
on time to first line chemotherapy failure controlling for visceral involvement a persistent non-
significant association was noted (figure 7). Despite the intention to control for other variables, due
to the lack of available observations and the fact that only concomitant visceral involvement was
significantly associated with first-line chemotherapy failure, only this covariate was included in the
model.
Figure 8 – Time to first-line chemotherapy failure according to menopausal status. Adjusted HR
controlling for visceral involvement is shown. HR – Hazard ratio; CI – Confidence interval; mo –
months; n – number.
Pre-menopause Post-menop.
Median time, mo. 15.25 17.40
Events, n (%) 8 (100) 22 (88.0)
Log-rank p = 0.399
Adjusted HR 0.72, 95% CI 0.29 – 2.77; p = 0.469
Page 26 of 38
Table 5 – Univariate association of baseline characteristics with time to first line chemotherapy
failure.
P-value
Menopausal status 0.399
Age at diagnosis of metastatic disease 0.673
Time from diagnosis to metastatic disease 0.272
Histology 0.882
Estrogen receptor status 0.867
HER2 status 0.915
Stage IV at diagnosis 0.418
Visceral involvement at diagnosis of metastatic disease
0.015
Performance status 0.240
d) Overall survival
When evaluating overall survival, the full cohort presented a median follow-up time of
approximately 49.6 months (IQR 28.4 – 60.4). A total of 34 events (85.0%) were documented, of
which 8 (100%) in the premenopausal group and 26 (81.3%) in the post-menopausal group. The
median time to death was of 30.9 months for pre-menopausal women and 52.8 for post-
menopausal women. In figure 8 a representation of the survival function according to menopausal
status is depicted. When evaluating the univariate association of menopausal status and other
clinical and pathological variables with change of first line chemotherapy, no significant associations
were found (table 6). However, HER status reached significance at 90% level of confidence, so we
decided to adjust menopausal status to this variable as well in a bivariate model. In this bivariate
model, no significant association was found as well (figure 8). As for the evaluation of time to first-
line chemotherapy failure, despite the intention to control for other variables, due to the lack of
available observations no other key variables were tested in the model.
Page 27 of 38
0
10
20
30
40
50
60
70
80
90
100
Pa
tients
aliv
e(%
)
20 18 14 6Post-menopause8 7 4 1Pre-menopause
No. at risk
0 20 40 60Follow-up time (months)
Pre-menopause Post-menopause
Figure 9 – Overall survival according to menopausal status. Adjusted HR controlling for HER2 status
is shown. HR – Hazard ratio; CI – Confidence interval; mo – months; n – number.
Table 6 – Univariate association of baseline characteristics with overall survival.
P-value
Menopausal status 0.221
Age at diagnosis of metastatic disease 0.999
Time from diagnosis to metastatic disease 0.294
Histology 0.733
Estrogen receptor status 0.618
HER2 status 0.094
Stage IV at diagnosis 0.662
Visceral involvement at diagnosis of metastatic disease
0.186
Performance status 0.377
Pre-menopause Post-menop.
Median time, mo. 30.89 52.82
Events, n (%) 8 (100) 26 (81.25)
Log-rank p = 0.221
Adjusted HR 0.63
95% CI 0.26 – 1.54; p = 0.317
Page 28 of 38
Discussion
Pre-clinical[30] and more recently clinical[31] evidence is accumulating in favor of an anti-cancer
activity of BPs, more specifically in the adjuvant treatment of post-menopausal women with breast
cancer. In the present study we intended to explore the differential anti-cancer effect of BPs
according to the hormonal environment in metastatic patients receiving ZA and chemotherapy. In
this subgroup of patients, no evidence of a differential anti-cancer activity of ZA between pre and
post-menopausal women was found when evaluating anti-cancer activity in terms of variation of
NTX (as surrogate for bone disease activity), variation of CA15.3 and CEA (as surrogate for overall
disease activity), time to first line chemotherapy failure (as surrogate for overall disease
progression) and overall survival.
BPs were developed and introduced in clinical practice as a tool to decrease the morbidity of bone
metastases given their properties as bone remodeling modulators, in specific as osteoclast
inhibitors. In this context, the primary endpoints of pivot trials were essentially centered on the rate
of SREs, time to first and subsequent SRE, improvement in pain control and general quality of life.
As a matter of fact, compared to placebo BPs reduced the SRE risk by 15% (RR 0.85, 95% CI 0.77-
0.94), with a median 28% reduction (range, 14–48%) of SREs.[41] However, no impact on overall
survival was found.[41]
Other studies tested the efficacy of BPs (zoledronic acid, pamidronate and clodronate) in the
metastatic setting with a focus on their anti-cancer activity with generally inconclusive findings.
However, based on the data derived from adjuvant studies, ours was the first study to test the
impact of menopausal status in the anticancer activity of BPs.
Two studies tested the anti-cancer activity of ZA in the metastatic setting. The MACS1295 study
(NCT01129336), previously referred as Z-ACT, tested the anti-tumoral action of ZA in patients with
HER2-negative metastatic breast cancer receiving standard first or second line therapy irrespective
of the menopausal status. In this study, patients were allocated to two different arms according to
the presence of bone metastases. Patients with bone metastases received standard therapy and ZA
(4 mg IV monthly during months 1-18), while patients without bone metastases were randomized
to receive standard therapy with or without ZA (same schedule). Primary outcome was the number
of participants with progression free survival at month 18, and secondary outcomes included the
percentage of patients with circulating tumor cells count above 5 cells per 7.5 mL of peripheral
blood, at several time points. Unfortunately, from the 150 planned patients without bone
metastases to be recruited, only 15 were included, making the interpretation of results not viable
(results available only at clinicaltrials.gov). Despite not stratifying according to the menopausal
status, this study, if properly completed, could have contributed to the improved understanding of
the clinical anti-tumor action of the zoledronic acid in the metastatic setting. Furthermore, the
BISMARK study (NCT01129336) indirectly tested the anti-tumor action of zoledronic acid.[42] In this
phase III non-inferiority study, 289 patients with bone metastases from breast cancer were
Page 29 of 38
randomized to receive zoledronic acid, either as a standard fixed schedule (4mg IV over 15 minutes
once every 3-4 weeks for 24 months) or under a set of rules dependent on NTX variation. Despite
the accrual limitations that led to premature study closure, the NTX-directed schedule group had a
higher proportion of patients presenting with an SRE, thus failing to demonstrate non-inferiority of
this approach. Of note, overall survival was selected as a secondary outcome in this study. Given
that patients in the standard schedule arm had more than the double of ZA administrations, this
study can help us gain some insight on the anti-tumor action of zoledronic acid in terms of overall
survival. No stratification based on menopausal status is planned so far.
Clodronate and pamidronate (in its oral formulation) were also tested as anti-neoplastic agents, in
specific, about their ability to prevent the development of bone metastases in patients with breast
cancer without clinical evidence of bone involvement and irrespective of menopausal status.[43–
45] Despite the fact that all studies were considerably underpowered to show a difference between
groups, when compared to placebo, no difference was found in the incidence of new bone
metastases. A recent meta-analysis summarized these findings, and found no overall effect of these
agents on the incidence of bone metastases in patients with metastatic breast cancer without
clinical evidence of bone metastases (HR 0.99, 95% CI 0.67 – 1.47).[41]
So far, while the studies performed in the early setting seem to show a protective action of BPs in
terms of disease recurrence and survival in patients that are estrogen deprived (post-menopausal
or under ovarian suppression), no such evidence is available in the metastatic setting. As a matter
of fact, the role of menopausal status is still unexplored in the metastatic setting; however,
fundamental differences between the early and advanced disease settings might limit the anti-
tumoral action of BPs in latter stages. First, the burden of disease of metastatic disease is larger,
than e.g., residual disease after treatment with curative intent. The intensity of BPs anti-tumor
effect might only be clinically relevant for smaller volumes of disease, hence limiting the action in
advanced cancer. Second, the central action of BPs might be mostly relevant before metastatic
disease fully develops. In this regard, BPs seem to interfere with the vascularization of the
premetastatic niche at several tentative distant sites of metastatic development.[30] Bone medulla
is also frequently considered a sanctuary for disseminated tumor cells.[46] Similarly to the action at
premetastatic niches, BPs might change bone medulla in such a way that hampers DTC survival,
either through the anti-angiogenic action or through the action on bone cells.[30] Third, post-
menopausal women present a more active bone metabolism, which some hypothesized that might
contribute to the establishment of bone metastasis.[18] BPs block bone resorption and unbalance
bone metabolism towards bone creation. This phenomena might impair the fixation of tumor cells,
a mechanism not transposable to the metastatic setting. On the other hand, certain mechanisms
may play similar roles both in the early and metastatic settings. BPs disrupt the vicious cycle
between cancer cells and bone remodeling activation, which decreases the availability of growth
factors entrapped in the bone matrix. Likewise, the immune activation, namely through γδ T cells,
would reasonably act on both settings of disease evolution.
Page 30 of 38
The present study presents important limitations which is compatible with its exploratory nature
and budget constraints. The first main limitation refers to the small sample size. Despite the
preplanned design as an exploratory study, the group of premenopausal women was considerably
small, which might not be representative of a conventional cohort of pre-menopausal patients.
Second, a relevant lost to follow-up was found at 9 months. Regardless of the intention to perform
the analysis of primary outcomes at 3 months, secondary outcomes interpretation is limited due to
this fact. Indeed, the late acceleration of urinary NTX in the pre-menopausal patients cannot be
properly valued in this context. Third, time to first line chemotherapy failure after the diagnosis of
metastatic disease was considerably long. In fact, several of these patients stopped first line
chemotherapy after a good clinical response and had a switch to hormone therapy before switching
to second line chemotherapy, as common practice in patients with hormone receptor positive
disease. The present analysis does not account for the type of hormone therapy received in the
transition to second line chemotherapy and this fact might have been a source of heterogeneity. As
a matter of fact, hormone therapy is known to interfere in bone remodeling[47, 48], and different
options of hormone therapy might change bone remodeling differently. Finally, the same threshold
of NTX normality was used both for pre and post-menopausal women for comparability reasons. It
is long known that several factors interfere with bone remodeling markers levels, such as age,
hormonal environment and gender. The persistently higher levels of NTX in post-menopausal
women at baseline and after the introduction of zoledronic acid and chemotherapy (despite all of
which non-significant) might be exclusively associated with this fact.
Noteworthy, to our best knowledge, none of the published studies so far tested the menopausal
status as a modulator of BPs action as an anti-tumor agent. In light of our findings and study
limitations this question remains unanswered.
Page 31 of 38
Conclusions
In a cohort of women with breast cancer and bone metastases receiving first line chemotherapy and
zoledronic acid, no difference seems to exist in terms of urinary NTX variation at 3 months according
to menopausal status. Furthermore, no differences was found in terms of CA15.3 and CEA variation
according to menopausal status. Finally, time to first line chemotherapy failure or overall survival
seems to be similar when evaluated according to menopausal status.
Taken together, these results do not support a differential action of zoledronic acid in the control of
bone metabolism (as a surrogate of cancer cells activity) or other cancer control related outcomes
according to menopausal status. However, these findings need to be considered side with the
discussed study limitations. As a matter of fact, this body of data can only be used as preliminary
findings for future studies intending to test the anti-cancer activity of BPs according to the hormonal
environment.
Page 32 of 38
Acknowledgments
I would like to acknowledge Doctor Irina Duarte for her decisively contribution to this work with the
quantification of NTX in the lab. I would also like to acknowledge Dr. Inês Vendrell for helping in the
collection of clinical information and all the personal support to complete the current Master.
Likewise, I would like to thank Irene Ferreira for the construction of figure 1 and all the enthusiasm
she always gave me on pursuing this Master. Furthermore, I would like to recognize the profound
contribution of Dr. Inês Vaz-Luís for all the mentorship around the development of this project since
its inception, as well as for all the contributions to my personal and professional development.
Moreover, I would like to thank Prof. Doctor Luís Costa for the personal inspiration and the role
model he became during my medical school years to aspire on being a physician scientist and more
recently on raising my awareness for the topic of bone metastases. He was also a cornerstone on
recruiting most of the patients analyzed. He further contributed with insightful comments to this
work and thesis. Finally, I would like to thank my family and friends, in special my mother, sisters
and father for the unconditional support in all aspects of my personal and professional life that
allowed me to complete this step in my academic development.
Page 33 of 38
List of abbreviations
ALP – Alkaline phosphatase
BALP – Bone specific alkaline phosphatase
BC/CM – Breast cancer/Cancro da mama
BP – Bisphosphonates
BSP – Non collagenous bone matrix proteins: bone sialoprotein
CA15.3 – Cancer antigen 15.3
CEA – Carcinoembrionic antigen
CI – Confidence interval
CT/QT – Chemotherapy/Quimioterapia
CTX – Carboxy-terminal crosslinked telopeptide of collagen type I
CXCL12 – chemokine (C-X-C motif) ligand 12
CXCR4 – chemokine (C-X-C motif) receptor 4
CXCR7 – chemokine (C-X-C motif) receptor 7
DKK-1 – Dickkopf-1
DTC – Dessiminated tumor cells
DPD – Deoxypyridoline
ER – Estrogen receptor
HER – human epidermal growth factor receptor 2
HR – Hazard ratio
Hyl – Hydroxylysine
Hyp – Hydroxyproline
ICTP – Carboxy-terminal crosslinked telopeptide of type I collagen
MT – Marcador tumoral
NTX – Amino-terminal crosslinked telopeptide of collagen type I
OC – Osteocalcin
OP – Osteopontin
OPG – Osteoprotegerin
P1NP – Propeptides of type I procollagen
PTH-rp – Parathyroid hormone related peptide
PoM – Post-menopausal
PrM – Pre-menopausal
PYD – Pyridinoline
RANK – Receptor activator of NFƙB
RANKL – Receptor activator of NFƙB ligand
SRE – Skeletal related event
TRAP5b – Tartarate resistant acid phosphatase 5b
TRAIL – TNF-related apoptosis-inducing ligand
USA – United States of America
ZA/AZ – Zoledronic acid/Ácido zoledrónico
Page 34 of 38
References
1. Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, et al. (2013) Cancer incidence and mortality patterns in Europe: Estimates for 40 countries in 2012. Eur J Cancer 49:1374–1403. doi: 10.1016/j.ejca.2012.12.027
2. Ferlay J, Soerjomataram I, Dikshit R, et al. (2015) Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 136:E359–86. doi: 10.1002/ijc.29210
3. National Cancer Institute (2014) Surveillance, Epidemiology, and End Results (SEER) Program. In: SEER Stat Fact Sheets Breast Cancer. http://seer.cancer.gov/.
4. Clarke M, Collins R, Darby S, Davies C, Evans V, Godwin J, Gray R, McGale P, Peto R, Wang Y.Clarke M, Collins R, Darby S, Davies C, Evans V, Godwin J, Gray R, McGale P, Peto R WY (2005) Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet 365:1687–717. doi: 10.1016/S0140-6736(05)66544-0
5. Jensen A, Jacobsen J, Norgaard M, et al. (2011) Incidence of bone metastases and skeletal-related events in breast cancer patients: A population-based cohort study in Denmark. BMC Cancer 11:29.
6. Zhang XH-F, Wang Q, Gerald W, et al. (2009) Latent bone metastasis in breast cancer tied to Src-dependent survival signals. Cancer Cell 16:67–78. doi: 10.1016/j.ccr.2009.05.017
7. Rosa Mendoza ES, Moreno E, Caguioa PB (2013) Predictors of early distant metastasis in women with breast cancer. J Cancer Res Clin Oncol 139:645–652. doi: 10.1007/s00432-012-1367-z
8. Ishizawar R, Parsons SJ (2004) c-Src and cooperating partners in human cancer. Cancer Cell 6:209–14. doi: 10.1016/j.ccr.2004.09.001
9. Lee JJ, Lotze MT (2009) Molecular basis of metastasis. N Engl J Med 360:1679; author reply 1679–1680. doi: 10.1056/NEJMra0805239
10. Roato I, Ferracini R (2013) Solid Tumours Show Osteotropism: Mechanisms of Bone Metastases. Clin Rev Bone Miner Metab 11:87–93. doi: 10.1007/s12018-013-9144-3
11. Mundy GR (1997) Malignancy and the skeleton. Horm Metab Res 29:120–127. doi: 10.1055/s-2007-979004
Page 35 of 38
12. Tranquilli Leali P, Doria C, Zachos A, et al. (2009) Bone fragility: Current reviews and clinical features. Clin Cases Miner Bone Metab 6:109–113.
13. Lipton A, Theriault RL, Hortobagyi GN, et al. (2000) Pamidronate prevents skeletal complications and is effective palliative treatment in women with breast carcinoma and osteolytic bone metastases: long term follow-up of two randomized, placebo-controlled trials. Cancer 88:1082–1090.
14. Coleman RE (2006) Clinical features of metastatic bone disease and risk of skeletal morbidity. Clin Cancer Res 12:6243s–6249s. doi: 10.1158/1078-0432.CCR-06-0931
15. Lipton A, Costa L, Coleman RE (2011) Bone turnover markers: tools for prognosis and monitoring response to bisphosphonates? Breast Dis 33:59–69. doi: 10.3233/BD-2010-0327
16. Costa L, Demers LM, Gouveia-Oliveira A, et al. (2002) Prospective evaluation of the peptide-bound collagen type I cross-links N-telopeptide and C-telopeptide in predicting bone metastases status. J Clin Oncol 20:850–856. doi: 10.1200/JCO.20.3.850
17. Wada N, Fujisaki M, Ishii S, et al. (2001) Evaluation of bone metabolic markers in breast cancer with bone metastasis. Breast Cancer 8:131–137. doi: 10.1007/BF02967492
18. Lipton A, Chapman J-AW, Demers L, et al. (2011) Elevated bone turnover predicts for bone metastasis in postmenopausal breast cancer: results of NCIC CTG MA.14. J Clin Oncol 29:3605–10. doi: 10.1200/JCO.2010.31.5069
19. Body JJ, Dumon JC, Gineyts E, Delmas PD (1997) Comparative evaluation of markers of bone resorption in patients with breast cancer-induced osteolysis before and after bisphosphonate therapy. Br J Cancer 75:408–412.
20. Demers LM, Costa L, Chinchilli VM, et al. (1995) Biochemical markers of bone turnover in patients with metastatic bone disease. Clin Chem 41:1489–94.
21. Brown JE, Cook RJ, Major P, et al. (2005) Bone turnover markers as predictors of skeletal complications in prostate cancer, lung cancer, and other solid tumors. J Natl Cancer Inst 97:59–69. doi: 10.1093/jnci/dji002
22. Coleman RE, Major P, Lipton A, et al. (2005) Predictive value of bone resorption and formation markers in cancer patients with bone metastases receiving the bisphosphonate zoledronic acid. J Clin Oncol 23:4925–4935. doi: 10.1200/JCO.2005.06.091
Page 36 of 38
23. Coleman R (2012) Randomized trial of marker-directed versus standard schedule zoledronic acid for bone metastases from breast cancer. J Clin Oncol 30:9–10.
24. Stopeck AT, Lipton A, Body J-J, et al. (2010) Denosumab compared with zoledronic acid for the treatment of bone metastases in patients with advanced breast cancer: a randomized, double-blind study. J Clin Oncol 28:5132–5139. doi: 10.1200/JCO.2010.29.7101
25. Coleman R, Body JJ, Aapro M, et al. (2014) Bone health in cancer patients: ESMO Clinical Practice Guidelines. Ann Oncol 25 Suppl 3:iii124–37. doi: 10.1093/annonc/mdu103
26. Buhaescu I, Izzedine H (2007) Mevalonate pathway: a review of clinical and therapeutical implications. Clin Biochem 40:575–84. doi: 10.1016/j.clinbiochem.2007.03.016
27. Rosen LS, Gordon D, Kaminski M, et al. (2003) Long-term efficacy and safety of zoledronic acid compared with pamidronate disodium in the treatment of skeletal complications in patients with advanced multiple myeloma or breast carcinoma: A randomized, double-blind, multicenter, comparative trial. Cancer 98:1735–1744. doi: 10.1002/cncr.11701
28. Sun L, Yu S (2013) Efficacy and safety of denosumab versus zoledronic acid in patients with bone metastases: a systematic review and meta-analysis. Am J Clin Oncol 36:399–403. doi: 10.1097/COC.0b013e31824be20e
29. Van Poznak CH, Von Roenn JH, Temin S (2011) American society of clinical oncology clinical practice guideline update: recommendations on the role of bone-modifying agents in metastatic breast cancer. J Oncol Pract 7:117–121.
30. Clézardin P (2013) Mechanisms of action of bisphosphonates in oncology: a scientific concept evolving from antiresorptive to anticancer activities. Bonekey Rep 2:267. doi: 10.1038/bonekey.2013.1
31. Coleman R, Gnant M, Paterson A, et al. (2014) Abstract S4-07: Effects of bisphosphonate treatment on recurrence and cause-specific mortality in women with early breast cancer: A meta-analysis of individual patient data from randomised trials. Cancer Res 73:S4–07–S4–07. doi: 10.1158/0008-5472.SABCS13-S4-07
32. Gralow JR, Biermann JS, Farooki A, et al. (2013) NCCN Task Force Report: Bone Health In Cancer Care. J Natl Compr Canc Netw 11 Suppl 3:S1–50; quiz S51.
Page 37 of 38
33. Gnant M, Mlineritsch B, Schippinger W, et al. (2009) Endocrine Therapy plus Zoledronic Acid in Premenopausal Breast Cancer. N Engl J Med 360:679–691. doi: 10.1056/NEJMoa0806285
34. Gnant M, Mlineritsch B, Stoeger H, et al. (2014) Zoledronic acid combined with adjuvant endocrine therapy of tamoxifen versus anastrozol plus ovarian function suppression in premenopausal early breast cancer: final analysis of the Austrian Breast and Colorectal Cancer Study Group Trial 12. Ann Oncol 26:313–320. doi: 10.1093/annonc/mdu544
35. Brufsky AM, Harker WG, Beck JT, et al. (2012) Final 5-year results of Z-FAST trial: Adjuvant zoledronic acid maintains bone mass in postmenopausal breast cancer patients receiving letrozole. Cancer 118:1192–1201. doi: 10.1002/cncr.26313
36. Coleman R, De Boer R, Eidtmann H, et al. (2013) Zoledronic acid (zoledronate) for postmenopausal women with early breast cancer receiving adjuvant letrozole (ZO-FAST study): Final 60-month results. Ann Oncol 24:398–405. doi: 10.1093/annonc/mds277
37. Coleman RE, Marshall H, Cameron D, et al. (2011) Breast-cancer adjuvant therapy with zoledronic acid. N Engl J Med 365:1396–405. doi: 10.1056/NEJMoa1105195
38. Paterson AHG, Anderson SJ, Lembersky BC, et al. (2012) Oral clodronate for adjuvant treatment of operable breast cancer (National Surgical Adjuvant Breast and Bowel Project protocol B-34): a multicentre, placebo-controlled, randomised trial. Lancet Oncol 13:734–42. doi: 10.1016/S1470-2045(12)70226-7
39. Kokufu I, Kohno N, Yamamoto M, Takao S (2010) Adjuvant pamidronate therapy prevents the development of bone metastases in breast cancer patients with four or more positive nodes. Oncol Lett 1:247–252. doi: 10.3892/ol_00000044
40. Jung J, Hwang G, Lee Y, et al. (2005) Pamidronate as adjuvant treatment for prevention of bone metastasis in breast cancer. ASCO Meet Abstr 23:888.
41. Wong MH, Stockler MR, Pavlakis N (2012) Bisphosphonates and other bone agents for breast cancer. Cochrane Database Syst Rev 2:CD003474. doi: 10.1002/14651858.CD003474.pub3
42. Coleman RRE Randomized trial of marker-directed versus standard schedule zoledronic acid for bone metastases from breast cancer. ASCO. p J Clin Oncol 30, 2012 (suppl; abstr 511)
Page 38 of 38
43. Kanis JA, Powles T, Paterson AHG, et al. (1996) Clodronate decreases the frequency of skeletal metastases in women with breast cancer. Bone 19:663–667. doi: 10.1016/S8756-3282(96)00285-2
44. Van Holten-Verzantvoort a. TM, Hermans J, Beex LV a M, et al. (1996) Does supportive pamidronate treatment prevent or delay the first manifestation of bone metastases in breast cancer patients? Eur J Cancer 32:450–454. doi: 10.1016/0959-8049(95)00564-1
45. Mardiak J, Bohunický L, Chovanec J, et al. (2000) Adjuvant clodronate therapy in patients with locally advanced breast cancer--long term results of a double blind randomized trial. Slovak Clodronate Collaborative Group. Neoplasma 47:177–80.
46. Weilbaecher KN, Guise TA, McCauley LK (2011) Cancer to bone: a fatal attraction. Nat Rev Cancer 11:411–425. doi: 10.1038/nrc3055
47. Vehmanen L, Elomaa I, Blomqvist C, Saarto T (2006) Tamoxifen treatment after adjuvant chemotherapy has opposite effects on bone mineral density in premenopausal patients depending on menstrual status. J Clin Oncol 24:675–680. doi: 10.1200/JCO.2005.02.3515
48. Eastell R, Adams JE, Coleman RE, et al. (2008) Effect of anastrozole on bone mineral density: 5-year results from the anastrozole, tamoxifen, alone or in combination trial 18233230. J Clin Oncol 26:1051–1058. doi: 10.1200/JCO.2007.11.0726